January 31, 2014 | Author: Mircea Popescu

mircea_popescu There. Let's discuss it here, I'll publish it afterwards.
herbijudlestoids Kthx.

mircea_popescu Ok so here we go. Maddoff's balancesheet, before the collapse, showed some assets. Right ?
herbijudlestoids Nominally, yes, but it was easy for anyone who checked to see that the maths of the assets was completely bogus, and multiple fiduciaries did this and reported to the SEC (which dutifully ignored).

mircea_popescu But let me make the argument for what it is. Those assets that he had, they were in fact somehow valued. On one approach to valuation, they came out X. On another approach to valuation, they came out Y. There's little convergence between the series of Xs and Ys. These are all facts, these all happened. Now. What is the root cause ? How do we know they don't happen again ?
herbijudlestoids Well I have an answer, pretty straightforward and based in science.

mircea_popescu Aha ?
herbijudlestoids Let's say I take the universe of hedge funds and I also take the Fama/French asset factors (value, momentum, liquidity, smallcap, etc), and do a factor decomposition against them. I can see which factors explain the returns of which funds.

mircea_popescu No argument at all.
herbijudlestoids Even though I don't know exactly which assets they might be invested in.

mircea_popescu What if your unit of account is broken ? What if in fact, unbeknowst to you, all this is a pyramid scheme based on trading baseball cards ?
herbijudlestoids The factors are self funded, they do not use the unit of account such as USD. It's long one asset, short another. E.g. long the 10 cheapest and short the 10 most expensive stocks in an index.

mircea_popescu Right... So what if it's zeroes everywhere. Does your model remain coherent ? Ie, 10 cheapest are all really worth 0. 10 most expensive are all really worth... 0.
herbijudlestoids The actual value is not under consideration. What I am trying to do is determine where the returns posted by the fund come from. If I can determine, for example, they come 90% value and 10% liquidity factor, then I know probably what the fund is up to.

mircea_popescu Yes, but you base this on some presumptions. Which is what I'm trying to illustrate to you.
herbijudlestoids If the returns arent explained by any factor then I know the fund needs exploration. Since most funds returns are easily explained by common factors, I know that most funds are not Madoff funds.

mircea_popescu Only within the context in which your presumptions hold true. This is the problem here : it's not that you can't build models that measure things. It's that it's in principle impossible to know if you're measuring everything you need to.
herbijudlestoids The factors are proxies, right?

mircea_popescu I'm not sure what you're asking there ?
herbijudlestoids They're crude approximations of what one might do to capture a perceived given premium. You're saying I presume various things, but I really don't, except on the most relative basis. I'm not presuming that expensive stocks are worth $1 and cheap ones are worth $2, just that on a relative basis one is cheaper than the other.

mircea_popescu Ok, want to run me through a model and we can evaluate this claim ?
herbijudlestoids Sure, here is a simple one which has had good returns : take the universe of stocks, for example wilshire 5000.

mircea_popescu So taken.
herbijudlestoids Figure out the earnings of the avg stock and also the total dollar volume for the avg stock. Figure out the TDV/earnings ratio for the avg stock. Weight investments inversely to this proportion, so stocks that have a high TDV/earnings ratio receive low investment, low TDV/earnings ratio receive high investment. That is the "model". Alternatively to self fund I can short the high TDV/earnings and long the low TDV/earnings ratio.

mircea_popescu Ok. So what does this model presume ?
herbijudlestoids Basically it presumes only that low liquidity stocks as measured by the TDV/earnings ratio will outperform stocks with a high TDV/earnings ratio. So lets say I calculate the return path over the last 30 years for such a strategy, and compare it using linear regression to the return of all hedge funds, I can see which funds have used a strategy like this to earn their returns. I know emphatically they are not Madoff funds.

mircea_popescu But first and foremost you must be able to calculate a TDV. You must also be able to obtain earnings figures.
herbijudlestoids Yes, both your statements are correct. I need information about the assets im investing in to inform the decision.

mircea_popescu Well you see ? And now to your comparison : you know which funds have DECLARED to not be Maddoff funds. What if all the funds are in fact lieing on their return sheets ? You will get a selection of funds which randomly happen to meet a criteria you picked.
herbijudlestoids Then the information is bad and I've made an uninformed investment, failing my job for due diligence.

mircea_popescu Ok. Now we're on the same page here. The original argument, re the boat that has sailed, is fundamentally exactly this.
herbijudlestoids What, that all the hedge funds are lying on their SEC filings?

mircea_popescu Which is to say : that there is an economic reality, such as it is. That it is the job of everyone, from the lowliest accountant to the cleverest CEO, to try and follow this reality. And that people being faillible, they do in fact make mistakes. They have in the past made mistakes. We will in the future discover they have made mistakes today. Now let's leave this for a moment and consider the problem of the Atlantic Investor. Suppose you have a thousand kilograms of gold, in 1750.
herbijudlestoids So supposed ;)

mircea_popescu For these you expect returns, and you are immortal. If you invest it all in Europe, you shall have in 1850 a return, such as it is, whatever's left past Talleyrandi. If you invest it all in the US, you shall in 1850 have a different return. This fundamental difference the Atlantic made was NOT obvious at the time. People thought the Louisiana purchase was a fair deal, at the time it was made, and the Alaska similarly.
herbijudlestoids Sure, but if I waited until 1751 to invest in the US or even 1800, long run returns show that the boat definitely hasn't sailed.

mircea_popescu Definitely. However! All markets have maturities. 10k bitcoin cost a pizza in 2010, 10k pizzas will cost a bitcoin in 2020.
herbijudlestoids Do you mean maturities in the actual financial sense of the term? I tried to calculate the maturity of bitcoin and couldn't.

mircea_popescu This understates a major problem of people. Major actors in 2014 were sitting in 2012 around a table having a beer. You being the guy that buys them a drink in 2012 is one thing. You being the guy trying to hire them in 2015 is another thing. (No, im discussing HR now. The financial maturity issue is insanity on bread, god help us.)
herbijudlestoids What youre describing is a reference point. Bitcoin becoming a reference point

mircea_popescu I think that may work, yes.
herbijudlestoids The way it works today is the USD is a reference point, we don't know how much things worth but all people carry in their head an approximate conversion table of the various price of things and by which determine the value of the USD.

mircea_popescu A business will forever be a business. It will obey the context it's in. But a country is a country, and while in 1750 you can easily acquire its dogged loyalty by a grant of 10k pounds, in 1850 you can't buy it for all the gold in the world. It will buy you. Like it or not.
herbijudlestoids And bitcoin is a country?

mircea_popescu Bitcoin is more of a country than it is a currency. Which, yes, sounds absurd.
herbijudlestoids No. That makes more sense to me than "bitcoin just is".

mircea_popescu That's unrelated, incidentally. That statement holds bitcoin as pars-pro-toto "cryptocurrency", a transcendent, if you will. This discussion actually discusses Bitcoin as a thing of this world.
herbijudlestoids Ok on the country analogy, I can kind of understand where you are coming from for the idea the boat has sailed. At some point the entity becomes sovereign.

mircea_popescu Exactly. At some point you can go to the kids and say "here, do it this way because I am older and wiser and I say so". At some later point you can not, if you haven't done it earlier. Because...well... now THEY are old enough.
herbijudlestoids And you feel that the point of sovereignety has come and gone.

mircea_popescu What I said was the time's ticking out, and by now it's probably gone by. Not in the sense that a perfect agent, Baby Jesus On A Stick, couldn't still do it, but in the sense that by now the barriers are so high and the average agent so dumb, that it won't likely happen anymore.
herbijudlestoids Ok, I disagree with nothing you have said.

mircea_popescu Well so at least we understand each other.
herbijudlestoids I guess the question is whether Bitcoin country is the US in 1750, or another shitty country that becomes sovereign and goes nowhere.

mircea_popescu I disclaim any absolute word-of-god bs, I have no idea that things are exactly this way because I'm fucking Moses or something. But I do think it's quite persuasive. And yes, the question may well be that, but that question has pretty good answers from other sources.
* herbijudlestoids tries to think of failed nation states.

mircea_popescu See, this is where the state analogy breaks down. Bitcoin is 1) incredibly attractive, such as the US never was in its history ; 2) incredibly permeable, such as again (even if only on the grouns of fighting disease and w/e) and 3) incredibly selective. Such as yet again. These three make it the best calender in history.
herbijudlestoids But it depends wholly on the Internet.

mircea_popescu It does not.
herbijudlestoids And a free internet at that.

mircea_popescu I could have my slaves run Bitcoin on their own skins, with sharp objects.
herbijudlestoids It doesnt?

mircea_popescu It strictly depends on math. No more. The only way to meaningfully attack Bitcoin is to break math.
herbijudlestoids For any convenient use, it depends on computers and the Internet.

mircea_popescu "Convenient" presumes the conclusion. "For any use including the Internet it requires the Internet". Was nobody and nothing convenient in 1960 ?
herbijudlestoids I'm not proposing this as a real world example, but just hear me out.

mircea_popescu Sure.
herbijudlestoids Let's say some APT like China or the US decides they don't like Bitcoin, for whatever reason, and they decide to fragment the Bitcoin network.

mircea_popescu Okay.
herbijudlestoids New peers are redirected to malicious nodes, and existing peers have their connections constantly reset (or similar).

mircea_popescu Wait a second. How are peers "redirected" ?
herbijudlestoids Well right now my understanding is that peers seed themselves to the network through the use of a DNS seed. Like in all p2p systems you don't just go around scanning random IPs looking for a friend. You seed to the network.

mircea_popescu Your attack model doesn't work in practice. So do you mean an APT doesn't like bitcoin or god doesn't like bitcoin ?
herbijudlestoids You think it's hard for an apt to disrupt a small handful of DNS seeds?

mircea_popescu It is impossible to disrupt the node finding as implemented. This is a lengthy technical discussion for which I don't really feel inclined (and perhaps am not the best to carry it with anyway).
herbijudlestoids Well ok, you seem to think new nodes can find peers. So lets just assume the APT is resetting all connections for bitcoin traffic.

mircea_popescu How does it identify bitcoin traffic ? I can send it as vacation pictures. So can you. Perhaps you currently are and don't even know.
herbijudlestoids You can but you don't and the large pools certainly don't.

mircea_popescu Not at the moment. Cheaper to fix that than to get a global resetting infrastructure in place.
herbijudlestoids I think judging from the great firewall of China is pretty clear if an APT wants to defreedom the Internet they can. Not that hard.

mircea_popescu On the same basis I think the contrary.
herbijudlestoids So why does Tor bother to have a slowly updating obfsproxy list?

mircea_popescu Tor is a complex topic (not least because it is a honeypot, not a legitimate service). Perhaps best left aside
herbijudlestoids I guess it will be hard for me to convince you Bitcoin depends on the Internet if you don't believe its possible to disrupt Bitcoin communication at any meaningful level.

mircea_popescu I am not saying it is not possible to disrupt it. Disruptions happen all the time. I run a number of nodes, they lose peers and gain new peers all the time. The idea is that it's impossible to control that disruption for less cost than it'd be to evade it. It's an economic problem, in the end. But that said, surely the
US could keep the consumer off Bitcoin.
herbijudlestoids But if there was an active attempt to say, convince your nodes that they are operating on a global network when they aren't...

mircea_popescu But how do you know which are my nodes ? Maybe I start a new, secret one, just for the hell of it.
herbijudlestoids I don't and I don't really care, the goal is just to fragment the pools to disrupt the network. I don't care whos nodes are whos, I just find the big pools and disrupt them.

mircea_popescu Tsk. Hashing doesn't depend on pools. Actually, according to most people studying it, pools are moreover bad.
herbijudlestoids I agree the pools are bad.

mircea_popescu In this sense, disrupting pools would strengthen the chain.
herbijudlestoids But that is the Bitcoin catch 22 : people centralise their infrastructure in the end anyway. Mostly for convenience.

mircea_popescu To some degree. Thats the key. They centralise to some degree. Anyway. Suppose tomorrow the Earth goes through a singularity, a bizarre quantum absurdity which renders all cables inoperable for communication. All of them. What happens now ?
herbijudlestoids Well im not denying that you can still 'bitcoin', but it will be much harder to do things like verify transactions :P

mircea_popescu No, I mean to the world. Clearly you can't get your shit off your camera, unless it has bluetooth. And you will need IR keyboards and monitors.
herbijudlestoids Oh well... I usually discount all scenarios mad max and above.

mircea_popescu Now, once people get a few months/years to accomodate to the new situation, do you imagine human society would be meaningfully different ?
herbijudlestoids Sure.

mircea_popescu Like how ?
herbijudlestoids Lower aggregate cortisol due to lack of cabling issues? :P

mircea_popescu Lol exactly. Same shit with Bitcoin. No internet ? It'll go over ham radio. People don't change just because some shit breaks. They just... don't. Once they discovered cocaine, cocaine was in constant use. Permanently. Once they discovered jacking off, they've been jacking off. The only difference between the two is that the latter being also older actually CHANGED political structure, societal mores and the very laws of the state under its unyielding pressure. "Cause it's there", says George, and right he is. It's there, and it never is going away.
herbijudlestoids I don't disagree with the idea in principle, but likely in such an event it would result in a new blockchain etc.

mircea_popescu Maybe it will. I'm not really arguing any detail here, it'd be absurd to try.
herbijudlestoids Ye.

———
  1. Charles Maurice de Talleyrand-Perigord. If you enjoy reading history, follow the life of this fellow, it's instructive. []
Category : Trilterviuri  | 4 responses.